The long-term objective of the proposed study is to determine the mechanism of nicotine dependence and smoking relapse. Smoking is well known to be associated with many human diseases, and thus causes a great burden on public health. Understanding why people smoke and why smokers can not quit smoking can provide a basis for useful strategies for planning health education and developing effective cessation programs for nicotine addicts. Consistent evidence has shown that genetic factors play a role in the maintenance of tobacco smoking. The final goal in smoking genetics researches should be to search for the loci that confer susceptibility to nicotine dependence. Although one recent report has found weak to moderate significance of linkages between several loci and cotinine level, without knowing the genetic model of the trait, linkage analysis is often not powerful in estimating the lods score. The best way to determine the genetic model and obtain the genetic parameters is to apply complex segregation analysis to family data. The authors propose to conduct secondary data analyses on very unique data generated in two smoking studies in our lab. The specific aims are to determine (1) whether there is familial aggregation in smoking and relapse; (2) whether there is a major gene effect responsible for the aggregation of each trait; (3) the best model for each trait, and (4) the persons genetically susceptible to nicotine addition, if there is a genetic effect. The probands were heavy smokers collected through smoking cessation programs. Data on approximately 500 families will be available for this study. Complex segregation analyses will be performed using S.A.G.E. program to fit the best model. Two phenotypes will be analyzed: smoking and smoking relapse. For smoking analysis, the phenotype for all probands will be affected, and for family members will be ever smoking or not. For relapse analysis, two classes of pheontypes will be defined for probands: relapse or not within 1 month and 1 year. The phenotype for family members will be relapse or not, without knowing their time to relapse. This is the first study, to our knowledge, to use complex segregation analysis for smoking behavior and relapse.